Hi All

i am doing join between two synonym using match tools. the first table has Customer Name , Balance and Amount. the second table has also Customer Name, Balance and Amount. both table hold different values for the measures. the issue when i am doing Join using Match Tools to generate a Report with one dimension and 4 measures it seems that i am reading same fields of measures of first table in the Match request. it should gives me different values for the 4 measures. i.e

Table1
-----------
Customer Name | Balance | Amount
john | 10 | 50
Maria | 20 | 70

Table2
-----------
Customer Name | Balance | Amount
john | 40 | 30
Maria | 60 | 80


Doing the Match where Customer in the By section as joining key and Balance,Amount in the SUM section.generate the report from the HOLD as
BY Customer and print details all the 4 measures gives this result:

Customer Name | Balance | Amount | Balance | Amount
john | 10 | 50 | 10 | 50
Maria | 20 | 70 | 20 | 70

i am looking for the following results:

HOLD
-----------
Customer Name | Balance | Amount | Balance | Amount
john | 10 | 50 | 40 | 30
Maria | 20 | 70 | 60 | 80



i appreciate your help and guidance
Qalqili

This message has been edited. Last edited by: Qalqili,

From the looks of your examples would the MORE command work better for you?

 TABLE FILE Table1
PRINT 
BALANCE
AMOUNT
BY CUSTOMER 
-*Where statement for Table1 here
ON TABLE PCHOLD FORMAT HTML
MORE
FILE Table2
-*Where statement for Table2 here
END

Actually, it looks to me that a JOIN would be the way to go as you need to put matching records from both tables in the same row.

JOIN CUSTOMER IN TABLE1 TO ALL CUSTOMER IN TABLE2

TABLE FILE TABLE1
PRINT TABLE1.CUSTOMER
      TABLE1.BALANCE  AS 'BALANCE_1'
      TABLE1.AMOUNT   AS 'AMOUNT_1'
      TABLE2.BALANCE  AS 'BALANCE_2'
      TBALE2.AMOUNT   AS 'AMOUNT_2'
END

If you need all data from TABLE1 even when no matching records exist in TABLE2, then you'll need a LEFT OUTER JOIN there to help you.

Now, if you need any records from either TABLE1 or TABLE2 regardless of the existence of matching information in each other table then you will definitely need to use MATCH OLD-OR-NEW logic.

In creating test data, I was able to get the desired results using the MATCH logic in the initial request:

  
TABLE FILE CAR
PRINT
COMPUTE CUST_NAME/A10 = ' ';
COMPUTE BALANCE/I5    = 0;
COMPUTE AMOUNT/I5     = 0;
BY COUNTRY NOPRINT
IF RECORDLIMIT EQ 1
ON TABLE HOLD AS TABLE1 FORMAT ALPHA
END
-RUN
HOLD AS TABLE2 FORMAT ALPHA

-WRITE TABLE1 JOHN         10   50 
-WRITE TABLE1 MARIA        20   70 
-CLOSE TABLE1
-RUN

-WRITE TABLE2 JOHN         40   30 
-WRITE TABLE2 MARIA        60   80 
-CLOSE TABLE2
-RUN

MATCH FILE TABLE1
SUM
BALANCE
AMOUNT
BY CUST_NAME
RUN
FILE TABLE2
SUM
BALANCE
AMOUNT
BY CUST_NAME
AFTER MATCH HOLD AS TEST3 OLD-OR-NEW
END
-RUN

TABLE FILE TEST3
PRINT *
END
-RUN

  
CUST_NAME BALANCE AMOUNT BALANCE AMOUNT 
JOHN      10      50     40      30 
MARIA     20      70     60      80 

Hi Mary

your code works well , i wounder is there a technique to diffrentiate the Measures from each other. i.e if i want to make a compute to take the difference from first Amount and first Balance and another compute to take the difference from secound Amount and second Balance from table2.

Regards
Qalqili

This message has been edited. Last edited by: Qalqili,

Sure there is (use the AS-phrase):

MATCH FILE TABLE1
SUM
BALANCE AS BALANCE1
AMOUNT AS AMOUNT1
BY CUST_NAME
RUN
FILE TABLE2
SUM
BALANCE AS BALANCE2
AMOUNT AS AMOUNT2
BY CUST_NAME
AFTER MATCH HOLD AS TEST3 OLD-OR-NEW
END

thansk very much Gamp, mary and every body for your help and assistance